vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

[SYCL] Fix WARP_SIZE=16 bug of Intel GPU (#8266)

Browse source 
* fix group_norm ut

* split softmax

* fix softmax

* add concat support condition

* revert debug code

* move QK_WARP_SIZE to presets.hpp
This commit is contained in:
luoyu-intel 2024-07-05 05:06:13 +00:00 • committed by GitHub
parent e235b267a2
commit a9554e20b6
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
8 changed files with 301 additions and 257 deletions
Download patch file Download diff file Expand all files Collapse all files

2
ggml/src/CMakeLists.txt
View file

@ -490,7 +490,7 @@ if (GGML_SYCL)
set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -fsycl-targets=nvptx64-nvidia-cuda")
add_compile_definitions(GGML_SYCL_WARP_SIZE=32)
else()
add_compile_definitions(GGML_SYCL_WARP_SIZE=32)
add_compile_definitions(GGML_SYCL_WARP_SIZE=16)
endif()
file(GLOB GGML_HEADERS_SYCL "ggml-sycl/*.hpp")